I believe these 2 professions need to be blended into a single occupation. Software engineering gets something tangible out the door, while data science is required to realize product features not possible through rules-based programming.

Software development is much less about “coding” than it is stitching together existing pieces, meaning the focus is shifting away from hardcore engineering towards crafting interesting product features.

The experimental nature of data science is more suited to how we need to be crafting software today. Mixing and matching components to discover the right product, as so many best practices have been “taken care of” through abstraction. This is inline with moving towards softer, more general skills.

IMO, neither the specialized data scientist nor the hardcore engineer is particularly suited to today’s software development practice. A hybrid Dev/DS role would arguably be a better skillset to bring into product development.